A single level cache has following specifications.
Access time = 2.5 ns
Line size = 64 bytes
Hit ratio = 0.95
Main memory uses a block transfer capability, and has first word (4 bytes) access time of 50 ns and access time for following words as 5 ns.
Compute the access time when there is a cache miss? Assume that the cache waits until the line has been fetched from main memory and then re-executes for a hit.
Will increasing the line size to 128 bytes, and a resulting increase in H (hit ratio) to .97, reduce the average memory access time?© BrainMass Inc. brainmass.com October 9, 2019, 8:17 pm ad1c9bdddf
Block size = Line size = 64 bytes = 64/4 or 16 (4 bytes) words
Time to fetch a line (or transfer a block) from main memory
= 50 ns to transfer first word + (15 * 5) ns to transfer remaining 15 words
= 50 + 75
= 125 ns
Cache access time on a cache miss
= time to completely fetch a line from ...
The solution gives step by step computations for better understanding of how to solve this kind of numericals.